Drawing Conclusions and Making Evidence-Based Claims

Section Drawing Conclusions and Making Evidence-Based Claims

The ability to look at data and draw reasonable conclusions is fundamental to scientific thinking. Students need to learn how to move from “Here’s what we found” to “Here’s what this means” while being appropriately confident about their claims.

🔗

Checkpoint 59.

Students survey 25 classmates about study habits and find that students who study with music report spending more time studying. A student concludes: “Music makes people study longer.” What’s problematic about this conclusion?

🔗

Hint.

Think about the difference between describing a pattern and explaining why it exists.

🔗

Solution.

The data shows a relationship but doesn’t prove causation. Students who study with music might study longer for other reasons (they enjoy studying more, they have different study habits, they’re more motivated), or students who plan to study longer might choose to use music. A better conclusion would be: “In our class, students who study with music report longer study times, but we need more information to understand why.”

🔗

Exploration 23. Try This Week: Evidence-Based Claim Practice.

Time needed: 20 minutes with any analyzed dataset

🔗

The Framework: Teach students to structure claims using this progression:

🔗

1. What we found: “Our data shows that...”

🔗

2. Our interpretation: “This suggests that...”

🔗

3. Our confidence level: “We’re [confident/somewhat confident/uncertain] because...”

🔗

4. What we’d need to know more: “To be more sure, we would need to...”

🔗

Elementary Example: “Our data shows that 18 out of 24 students prefer sunny days to rainy days. This suggests that most kids in our class like sunny weather better. We’re confident about this because we asked almost everyone. To be more sure, we would need to ask other classes and see if we get similar results.”

🔗

Secondary Example: “Our data shows that students who eat breakfast score an average of 8 points higher on morning quizzes. This suggests there might be a connection between breakfast and test performance. We’re somewhat confident because we saw this pattern consistently, but we would need to control for other factors like sleep, study time, and family income to be more sure it’s really about breakfast.”

🔗

Checkpoint 60.

Why is it important for students to use words like “might,” “suggests,” and “appears to” when making claims from data?

🔗

Hint.

Think about the difference between certainty and probability in data interpretation.

🔗

Solution.

Probabilistic language reflects the reality that data analysis involves uncertainty. Even strong patterns in data might not hold in other contexts, with different people, or under different conditions. Using tentative language keeps students intellectually honest and helps them avoid overconfident claims that can’t be supported by their evidence.

🔗

Checkpoint 61. Correlation vs. Causation Practice.

One of the most important interpretation skills is distinguishing between relationships and causes.

🔗

(a)

Students find that class periods with more questions from students tend to have higher test scores the next day. What can they reasonably conclude?

🔗

Answer.

There’s a relationship between student questions and test performance, but they can’t conclude that questions cause better scores.

🔗

(b)

What other factors might explain both more questions and higher test scores?

🔗

Answer.

Student engagement, lesson clarity, topic difficulty, class preparation, or teaching approach that day.

🔗

Teaching students to identify alternative explanations helps them think more critically about cause-and-effect claims.

🔗

Fun examples to help students distinguish between correlation and causation.

🔗

Checkpoint 62.

How can you help students learn to ask questions that can actually be answered with data?

🔗

Hint.

Think about the difference between questions that can be investigated and those that are purely matters of opinion.

🔗

Solution.

Students can practice by sorting questions into “Can be investigated with data” vs. “Matter of opinion/values.” Then practice transforming vague questions (“Are teenagers lazy?”) into investigable ones (“How do teenage sleep patterns compare to recommended amounts?”). This helps them formulate questions that can lead to meaningful conclusions.

🔗

Prev Top Next